Which would you rather remember: 184.108.40.206 or example.com?
Paul Mockapetris invented DNS in 1983 so that we wouldn’t have to memorize scores of random number sets to navigate the Internet. In the decades since, the world and the Internet have become increasingly dynamic, and applications have become distributed, causing DNS to play a larger part in business operations and application delivery.
In the modern ecosystem, it’s important to route users not just to a single application server, but to the best server, providing the highest quality of service for each request, and—here’s the kicker—all this must be done in real time.
Guest article by Kris Beevers, Cofounder and CEO, NS1
Long thought of as the Yellow Pages for the Internet, DNS is taking on a more prominent role as application architectures change and the Internet continues to evolve. Because of this, DNS is now more important than ever for developers, engineers, and operations teams who build modern applications and websites.
How DNS’ role is changing
Here’s why DNS is a critical aspect of Internet infrastructure: applications and websites become unavailable—essentially “invisible”—if a DNS service fails. This results in downtime and loss of revenue.
But DNS presents a powerful opportunity: it is the ingress point to most applications and websites. This means DNS lookup has the potential to be the first and most impactful chance to make a decision about which application endpoint should service a user.
Significant technology evolution has occurred in the last several years to enable distributed applications: automation tools, application frameworks, IaaS offerings, database technologies, increased bandwidth, and so on. However, DNS and traffic management technology has not evolved at the pace of applications themselves, and DNS-based routing decisions are often limited to “geographic routing.” This attempts to route users based on metrics (like distance) that may not actually take a user on an ideal path, and through a complex Internet subject to congestion, peering constraints, fiber cuts, malicious traffic, and other factors. Basically, until recently, we were in a state of “dumb DNS.”
Now, though, a new crop of intelligent DNS solutions goes far beyond the rote capabilities of traditional DNS, and can dramatically improve application performance and end-user experience.
On the network side, intelligent DNS can help solve a number of common application delivery problems by directing users to application endpoints that minimize packet loss, jitter, latency, and other factors.
On the infrastructure side, intelligent DNS helps route traffic around outages, optimize endpoint selection based on real-time telemetry, and even cloudburst to meet spikes in demand. These capabilities are driven by the ability to ingest infrastructure and network data, often specific to an application, and automatically make decisions based on that data in real time.
Is a modern DNS platform right for you?
Next-generation DNS solutions offer the intelligence and capabilities that traditional in-house and legacy solutions simply cannot offer, yielding greater speed, operational control, and reliability.
These features can be leveraged via SaaS-model managed DNS networks, or even in fully managed on-premise deployments for internal use cases like service discovery or corporate DNS. Increasingly, managed DNS and on-premise DNS are tightly coupled for single-pane-of- glass manageability, visibility, and automation of DNS and traffic management across a company’s infrastructure.
An intelligent DNS solution will include these four critical capabilities.
A global network
You need a deep technical background, real-time visibility, and some serious infrastructure to keep critical DNS systems up, respond to increasingly complex threats, and deliver high performance to a global audience. The expertise, resources and experience behind a managed DNS provider ensure a resilient and available system. Identifying providers that use multiple upstream carriers to prevent failure due to network outages can effectively manage threats and offer a global, anycasted network for optimal resiliency and speed.
To prevent any single node from getting overloaded and to ensure efficient use of your infrastructure, traffic should be spread across your application endpoints. Ask potential DNS providers what types of load balancing they are capable of – from simple round robin load balancing to advanced configurations including weighting, session affinity, and load shedding.
Providers of intelligent DNS use real-time data feeds and are able to integrate with third-party monitoring services or leverage built-in monitoring capabilities, to prevent downtime and route traffic based on data from multiple sources, including monitoring systems and performance metrics from real users.
Application-specific traffic management
Geographic and performance-based traffic routing features can help you scale or segment your users, or direct them based on fine-grained, real-time telemetry to the optimal application endpoint. Ask providers what types of routing filters and algorithms are available to help meet your traffic management needs, and look for tools that enable easy configuration of custom routing strategies.
A New Era in Managed DNS
Cloud-based services, the Internet of Things, widespread mobility, and rapidly evolving application architectures have transformed the digital landscape. Now, DNS technologies have evolved to keep pace, enabling reliable, fast, and truly intelligent traffic management in this ubiquitous, mission-critical stage of the application delivery lifecycle.
While in some cases there is still a place for simple traditional in-house DNS, affordable managed DNS services—both in the cloud and on-premise—now exist to help developers build secure, highly available, high-performance applications.
For more information, also see the related Aberdeen Group content, Bad DNS Trips Can Ruin Web Performance.
Kris Beevers is an Internet infrastructure geek and serial entrepreneur who’s started two companies, built the tech for two others, and specializes in architecting high-volume, globally distributed Internet infrastructure. Before founding NS1, Kris built CDN, cloud, bare metal, and other infrastructure products at Voxel, a NY-based hosting company that sold to Internap (NASDAQ:INAP) in 2011.